When tracing a person based on videos, what is really of interest is to trace a track of a target person, i.e., to obtain track information of a target person. In order to obtain track information of a target person, all videos containing the target person need to be obtained, and it is possible to determine the track information of the target person and implement the track tracing of a target person based on locations of image acquisition devices that acquire these videos.
However, in above methods, it is needed to analyze a large number of videos to analyze whether the large number of videos contain a target person, thus the analysis process is time-consuming, laborious, and very complex.